## HINT A Hierarchical Index for Intervals in Main Memory

Indexing intervals is a fundamental problem, finding a wide range of applications . Recent work on managing large collections of intervals in mainmemory focused on overlap joins and temporal aggregation problems . In thispaper, we propose novel and efficient in-memory indexing techniques for intervals .…

## A Prioritized Trajectory Planning Algorithm for Connected and Automated Vehicle Mandatory Lane Changes

We introduce a prioritized system-optimal algorithm for mandatory lane change(MLC) behavior of connected and automated vehicles (CAV) from a dedicated lane . Our approach outperforms the traditional gap acceptance model . Our experiments on synthetic data show that the proposed algorithm improves the traffic networkefficiency by attaining higher speeds in the dedicated lane and earlier MLCpositions while ensuring a low computational time .…

## TeraMIMO A Channel Simulator for Wideband Ultra Massive MIMO Terahertz Communications

There is a consensus among multiple research initiatives on the essential role of terahertz communications in the sixth generation of wireless communication systems and beyond . Aiming atcatalyzing THz communications research, we propose TeraMIMO, an accurate MATLABsimulator of statistical THz channels .…

## Understanding and Avoiding AI Failures A Practical Guide

As AI technologies increase in capability and ubiquity, AI accidents are becoming more common . By focusing on system properties near accidents instead of the root cause of accidents, we identify where attention should be paid to safety for current generation AI systems .…

## Fast Text Only Domain Adaptation of RNN Transducer Prediction Network

RNN-transducer models can be effectively adapted to new domains using only small amounts of textual data . By taking advantage of model’s inherent structure, where the prediction network is interpreted as a language model, we can apply fast adaptation to the model .…

## ManipulaTHOR A Framework for Visual Object Manipulation

The domain of Embodied AI has recently witnessed substantial progress in navigating agents within their environments . We propose a framework for object manipulation builtupon the physics-enabled, visually rich AI2-THOR framework . ArmPointNav offers new challenges including 3D obstacle avoidance, manipulating objects in thepresence of occlusion, and multi-object manipulation that necessitates longterm planning .…

## Restoring degraded speech via a modified diffusion model

DiffWave has shown state-of-the-artsynthesized speech quality and relatively shorter waveform generation times,with only a small set of parameters . We replace the mel-spectrum upsampler in DiffWave with a deep CNN upsamplator . The model is trained using the original speech waveform, but conditioned on the degraded speechmel-spectrums .…

## The Density Fingerprint of a Periodic Point Set

Fingerprinth is a fast algorithm based on Brillouin zones and related inclusion-exclusionformulae . We prove invariance under isometries,continuity, and completeness in the generic case . The proof of continuity integratesmethods from discrete geometry and lattice theory . We have implemented the algorithm and describe its application tocrystal structure prediction .…

## Real Time Trajectory Optimization in Robot Assisted Exercise and Rehabilitation

This work focuses on the optimization of the training trajectory orientationusing a robot as an advanced exercise machine (AEM) and muscle activations asbiofeedback . Muscle recruitment patterns depend on trajectory parameters of the AEMs and correlate with the efficiency of exercise .…

## On Buggy Resizing Libraries and Surprising Subtleties in FID Calculation

Fr\’echet Inception Distance (FID)score is widely used to evaluate generative models . Each FID implementation uses a different low-level image processing process . Image resizing functions in commonly-used deep learning libraries often introduce aliasing artifacts . We observe that numerous subtle choices need to be made for FID calculation .…

## Pose Controllable Talking Face Generation by Implicitly Modularized Audio Visual Representation

Previous methods rely on pre-estimated structural information such as landmarks and 3D parameters . However, the inaccuracy of such estimated information under extreme conditions would lead to degradation problems . In this paper, we propose a clean yet effective framework to generatepose-controllable talking faces .…

## Model Driven Deep Learning Based Channel Estimation and Feedback for Millimeter Wave Massive Hybrid MIMO Systems

This paper proposes a model-driven deep learning (MDDL)-based channelestimation and feedback scheme for wideband millimeter-wave (mmWave) massivehybrid multiple-input multiple- input multiple-output (MIMO) systems . The angle-delay domain channels’ sparsity is exploited for reducing the overhead for estimating the high-dimensional channels from a limited number of radio frequency (RF) chains at the base station .…

## Multi point Coordination in Massive MIMO Systems with Sectorized Antennas

Non-cooperative cellular massive MIMO, combined with power control, is knownto lead to significant improvements in per-user throughput compared with conventional LTE technology . We demonstrate that employing sectorizedantenna elements mitigates the detrimental effects of pilot contamination by rejecting a portion of interfering pilots in the spatial domain during channelestimation phase .…

## Misinformation Believability and Vaccine Acceptance Over 40 Countries Takeaways From the Initial Phase of The COVID 19 Infodemic

The COVID-19 pandemic has been damaging to the lives of people all around the world . Accompanied by the pandemic is an infodemic, an abundant anduncontrolled spreading of potentially harmful misinformation . This paper presentsfindings from a global survey on the extent of worldwide exposure to theCOVID-20 pandemic .…

## Identifying Actions for Sound Event Classification

In Psychology, actions are paramount for humans to perceive and separatesound events . We propose a new Psychology-inspired approach for SEC that includes identification of actions via human listeners . We used crowdsourcing to have listeners identify 20 actions that in isolation or in combination may have produced any of the 50 sound events in the well-studied dataset ESC-50 .…

## Topological Simplifications of Hypergraphs

We study hypergraph visualization via its topological simplification . In simplifying a hypergraph, we allowvertices to be combined if they belong to almost the same set of hyperedges . Our proposed approaches are general, mathematically justifiable, and they putvertex simplification and hyperedge simplification in a unifying framework.…

## Voice2Mesh Cross Modal 3D Face Model Generation from Voices

Previous works for cross-modal facesynthesis study image generation from voices . However, image synthesis includesvariations such as hairstyles, backgrounds, and facial textures, that arearguably irrelevant to voice . Weinstead investigate the ability to reconstruct 3D faces to concentrate on onlygeometry, which is more physiologically grounded .…

## Towards a more efficient approach for the satisfiability of two variable logic

We revisit the satisfiability problem for two-variable logic, known to be NEXP-complete . The upper bound is usually derived from its well known “exponential size model” property . We show that without the equality predicate SAT(FO2) is in factequivalent to CIS in succinct representation .…

## QCSP on Reflexive Tournaments

We give a complexity dichotomy for the Quantified Constraint SatisfactionProblem QCSP(H) when H is a reflexive tournament . We prove that if H has both its initialand final strongly connected component (possibly equal) of size 1, then QCSP is NP-hard .…

## Label Synchronous Speech to Text Alignment for ASR Using Forward and Backward Transformers

The speech-to-text alignment is a problem of splitting long audio recordings with un-aligned transcripts into utterance-wise pairs of speech and text . Unlike conventional methods, the proposed method re-defines the problem as a label-synchronous text mapping problem . Thisenables an accurate alignment benefiting from the strong inference ability of the state-of-the-art attention-based encoder-decoder models, which cannot be applied to the conventional methods .…

## Wireless Sensing With Deep Spectrogram Network and Primitive Based Autoregressive Hybrid Channel Model

Human motion recognition (HMR) based on wireless sensing is a low-costtechnique for scene understanding . Current HMR systems adopt support vectormachines (SVMs) and convolutional neural networks (CNNs) to classify radarsignals . On the other hand, training a machinelearning model requires a large dataset, but data gathering from experiment iscost-expensive and time-consuming .…

## Acyclic Star and Injective Colouring Bounding the Diameter

We examine the effect of bounding the diameter for well-studied variants of the Colouring problem . A colouring is acyclic, star, or injective if any twocolour classes induce a forest, star forest or disjoint union of vertices andedges . The corresponding decision problems are Acyclic Colouring,Star Colouring and Injective Colouring .…

## Efficient Sparse Coding using Hierarchical Riemannian Pursuit

Sparse coding is a class of unsupervised methods for learning a sparserepresentation of the input data in the form of a linear combination of adictionary and a sparse code . Initial non-convexapproaches learn the dictionary in the sparse coding problem sequentially in anatom-by-atom manner, which leads to a long execution time .…

## Three Dimensional Mesh Steganography and Steganalysis A Review

Three-dimensional (3-D) meshes are commonly used to represent virtualsurfaces and volumes . Over the past decade, 3-D meshes have emerged in industrial, medical, and entertainment applications . We propose a new taxonomy of steganographic algorithms with four categories: 1) two-state domain, 2- LSB domain, 3) permutation domain, 4) transform domain .…

## Uplink Performance Analysis of Cell Free mMIMO Systems under Channel Aging

In this paper, we investigate the impact of channel aging on the uplinkperformance of a cell-free~(CF) massive multiple multiple-input multiple- input multiple-output (mMIMO) system . We present a new model for the temporal evolution of the channel, which allows thechannel to age at different rates at different access points .…

## Model aided Deep Reinforcement Learning for Sample efficient UAV Trajectory Design in IoT Networks

Deep Reinforcement Learning (DRL) has become a prominent paradigm to designtrajectories for autonomous unmanned aerial vehicles . We propose a model-aided deep Q-learning approach that, in contrastto previous work, requires a minimum of expensive training data samples and isable to guide a flight-time restricted UAV on a data harvesting mission without prior knowledge of wireless channel characteristics and limited knowledge of node locations .…

## What s The Context Long Context NLM Adaptation for ASR Rescoring in Conversational Agents

Neural Language Models (NLM) consistently outperform n-gram language models and NLMs that use limited context . We propose the use of attention layer over lexicalmetadata to improve feature based augmentation . We adapt ourcontextual NLM towards user provided on-the-fly speech patterns by leveragingencodings from a large pre-trained masked language model and performing fusion with a Transformer-XL based NLM .…

## Stable Nonlinear and IQ Imbalance RF Fingerprint for Wireless OFDM Devices

An estimation method of Radio Frequency fingerprint (RFF) based on the physical hardware properties of the nonlinearity and in-phase and quadrature(IQ) imbalance of the transmitter is proposed for the authentication of wireless orthogonal frequency division multiplexing (OFDM) devices . The proposed RFfingerprinting method is helpful for the high-strength authentication of theOFDM communication devices with subtle differences from the same model and sameseries .…

## Towards Exploratory Landscape Analysis for Large scale Optimization A Dimensionality Reduction Framework

Little is known about the scalability of the ELA approach for large-scale optimization . Two important feature classes (ela_level and ela_meta) cannot be applied to optimization due to their high computational cost . Adimensionality reduction framework proposes a framework that computes features in a reducedlower-dimensional space than the original solution space .…

## Lossless Compression with Latent Variable Models

We develop a simple and elegant method for lossless compression using latentvariable models . The method involves interleaving encode and decode steps, andachieves an optimal rate when compressing batches of data . We demonstrate it firstly on the MNIST test set, showing that state-of-the-art losslesscompression is possible using a small variational autoencoder (VAE) model .…

## Room adaptive conditioning method for sound event classification in reverberant environments

One of the unpredictable and detrimental factors in performance, especially in indoor environments, is reverberation . We propose a conditioning method that provides room impulseresponse (RIR) information to help the network become less sensitive toenvironmental information and focus on classifying the desired sound .…

## On reduction and normalization in the computational core

We study the reduction in a lambda-calculus derived from Moggi’s computational core . The reduction relationconsists of rules obtained by orienting three monadic laws . We investigate the central notions of returning a value versus having a normal form, and address the question of normalizing strategies .…

## Voxel Structure based Mesh Reconstruction from a 3D Point Cloud

Mesh reconstruction from a 3D point cloud is an important topic in the field of computer graphic, computer vision, and multimedia analysis . In this paper, we propose a voxel structure-based mesh reconstruction framework . It providesthe intrinsic metric to improve the accuracy of local region detection .…

## A convergent numerical scheme for a model of liquid crystal dynamics subjected to an electric field

We present a convergent and constraint-preserving numerical discretization of a mathematical model for the dynamics of a liquid crystal subjected to anelectric field . This model can be derived from the Oseen-Frank director fieldtheory . We show that the method is stable even when singularities develop, and predictions about the alignment of the director field with the electric field are confirmed .…

## On the Width of Regular Classes of Finite Structures

In this work, we introduce the notion of decisional width of a finiterelational structure . We also introduce the idea of decisionality of a regular class offinite structures . Our main result states that given a first-order formula, and a finite automaton F over a suitablealphabet B, one can decide in time f (f) whether some {\tau}-structure in C satisfies {\psi}.…

## Discrete Vector Bundles with Connection and the Bianchi Identity

We develop a discrete theory of vector bundles with connection that isnatural with respect to appropriate mappings of the base space . The mainobjects are discrete vector bundle valued cochains . The central operators are adiscrete exterior covariant derivative and a combinatorial wedge product .…

## Carbon Emissions and Large Neural Network Training

The computation demand for machine learning (ML) has grown rapidly recently, which comes with a number of costs . We calculate the energy use and carbon footprint of several recent large models-T5, Meena, GShard, Switch Transformer, andGPT-3-and refine earlier estimates for the neural architecture search that found Evolved Transformer .…

## Public Perception of the German COVID 19 Contact Tracing App Corona Warn App

In Germany, the related app is called Corona-Warn-App, and by end of 2020, it had 22.8 million downloads . Contacttracing is a promising approach for containing the spread of the novelcoronavirus . It is only effective if there is a large user base, which brings new challenges like app users unfamiliar with using smartphones or apps .…

## On Sampling Based Training Criteria for Neural Language Modeling

In this work, we consider Monte Carlo sampling, importancesampling, a novel method we call compensated partial summation, and noisecontrastive estimation . The essence of these sampling methods is that the softmax-related traversalover the entire vocabulary can be simplified, giving speedups compared to the baseline .…

## Sensitivity as a Complexity Measure for Sequence Classification Tasks

The sensitivity of a function quantifies the number of disjoint subsets of the input sequence that can each be individually changed to change the output . We argue that standard sequence classification methods are biased towards learning low-sensitivity functions, so that tasks requiring high sensitivity are more difficult .…

## Bias Aware Loss for Training Image and Speech Quality Prediction Models from Multiple Datasets

Ground truth used for training image, video, or speech quality prediction models is based on the Mean Opinion Scores (MOS) obtained from subjectiveexperiments . Usually, it is necessary to conduct multiple experiments, mostly with different test participants, to obtain enough data to train quality models .…

## FD JCAS Techniques for mmWave HetNets Ginibre Point Process Modeling and Analysis

Co-design of full-duplex (FD) radio with jointcommunication and radar sensing (JCAS) techniques in millimeter-wave (mmWave)heterogeneous networks (HetNets) Spectral co-existence of radar and communication systems causes mutual interference between the two systems . Focusing on thedetection performance, we propose a cooperative detection technique, whichexploits the sensing information from multiple base stations (BSs) In real-world network scenarios, the locations of the BSsare spatially correlated, exhibiting a repulsive behavior .…

## Wadge Degrees of Classes of omega Regular k Partitions

We develop a theory of k-partitions of the set of infinite words recognizable by classes of finite automata . The theory enables to complete proofs of results about topological classifications of the (aperiodic)omega-regular k-Partitions and provides tools for dealing with other similar questions .…

## Game Theory to Study Interactions between Mobility Stakeholders

Increasing urbanization and exacerbation of sustainability goals threaten theoperational efficiency of current transportation systems . Rise of private, profit-maximizing Mobility Service Providers leveragingpublic resources, such as ride-hailing companies, entangles current regulationschemes . In this paper, we provide a game-theoretic framework to study interactions between stakeholders of the mobility ecosystem, modeling regulatory aspectssuch as taxes and public transport prices, as well as operational matters forMobility Services Providers such as pricing strategy, fleet sizing, and vehicledesign .…

## Improvement of Normal Estimation for PointClouds via Simplifying Surface Fitting

With the burst development of neural networks in recent years, the task of normal estimation has once again become a concern . Similar to the principle ofOccam’s razor, that is, the simpler is better. We observe that a moresimplified process of surface fitting can significantly improve the accuracy of the normal estimation .…

## Reduced order modeling of LPV systems in the Loewner framework

We propose a model reduction method for LPV systems . We consider LPV state-space representations with an affine dependence on the schedulingvariables . The main idea behind the proposed method is to compute the reducedorder model in such a manner that its frequency domain transfer function matches with that of the original model for some frequencies .…

## A note on some information theoretic divergences between Zeta distributions

In this short communication, we first report a closed-form formula forcalculating the $\alpha$-divergences and Sharma-Mittal divergences . Then we report a formula to calculate theKullback-Leibler divergence between any two zeta distributions . We illustratethese formula with examples .…

## Macroeconomic forecasting with statistically validated knowledge graphs

This study leverages narrative from global newspapers to construct theme-based knowledge graphs about world events . It shows that featuresextracted from such graphs improve forecasts of industrial production in threelarge economies compared to a number of benchmarks . The theme categories “disease” and “economic” have the strongest predictive power during the time period that we consider .…

## How to Identify and Authenticate Users in Massive Unsourced Random Access

In unsourced random access (U-RA) packets do not include any address field for the user, which maximizes thenumber of useful bits that are transmitted . This paper presents a schemethat builds upon an underlying U-RA protocol and solves the problem of user identification and authentication .…

## Predictive analytics using Social Big Data and machine learning

The ever-increase in the quality and quantity of data generated fromday-to-day businesses operations in conjunction with the continuously importedrelated social data have made the traditional statistical approaches inadequateto tackle such data floods . This chapter sheds the light on coreaspects that lay the foundations for social big data analytics .…